Estimation of Water Volume in Reservoirs Using Spectral Indices: A Case Study with NDVI in the Milagres Reservoir (Pernambuco, Brazil) (original) (raw)

The mapping of land use and land cover is essential for understanding changes in the terrestrial environment, identifying natural transformations, and human interventions over different periods and locations. However, traditional data collection and analysis methods for monitoring water bodies can be expensive and labor-intensive, especially in extensive areas. In this context, Remote Sensing stands out as an accessible technique that utilizes satellite images and drones to monitor the Earth. This study focuses on the use of the Normalized Difference Vegetation Index (NDVI) and Normalized Difference Water Index (NDWI) to monitor water bodies, with a specific focus on the Milagres Reservoir, part of the São Francisco River Integration Project with the Northeastern Basins (PISF) in Pernambuco, Brazil. NDVI is employed to distinguish aquatic areas, while NDWI values are explored to identify the presence of water. Landsat-8 satellite images were processed using the QGIS software, including atmospheric correction to enhance image quality. The obtained data were compared with information provided by the Ministry of Regional Development (MIDR), enabling the creation of a Water Level-Area-Volume (WAV) curve and the development of polynomial equations to estimate water volumes based on NDVI values. The results showed statistically acceptable correlations between estimated and real values, especially for water levels and areas. However, volume values exhibited larger discrepancies, suggesting the need for periodic reviews of the reservoir's bathymetry. It is recommended to explore images from other satellites, such as Sentinel-2A and CBERS 4A, to improve result accuracy. Additionally, the inclusion of NDWI in the methodology can enrich the analysis, providing a more comprehensive view of water resource distribution. In summary, this study offers a promising approach for remote water body monitoring, with the potential to enhance water resource management in regions like Brazil's Northeastern semi-arid area.